scheduleCudaTask

Fully qualified name: omni::graph::image::unstable::scheduleCudaTask

template<typename TParams>
inline void omni::graph::image::unstable::scheduleCudaTask(

omni::graph::core::GpuFoundationsInterfaces *gpu,

omni::usd::hydra::HydraRenderProduct *rp,

uint32_t deviceIndex,

ComputeParams<TParams> const &computeParams,

const char *renderOpName,

std::function<void(ComputeParams<TParams>*, cudaStream_t)> const &computeCuda,

)

Schedule a CUDA task on the post render graph.

Parameters:

• gpu – [in] The GPU interface.

• rp – [in] The render product on which the CUDA computation is applied.

• deviceIndex – [in] The device index on which the CUDA task is scheduled.

• computeParams – [in] The parameters of the computation.

• renderOpName – [in] The name of the render op in the render graph.

• computeCuda – [in] The CUDA computation entry point.

Template Parameters:

TParams – The concrete type of the CUDA compute params.